|
|||||
|
|
| 基于图像中心矩和特征向量的目标识别方法 | |
| 引用本文： | 王琪,万中南,韩俊伟.基于图像中心矩和特征向量的目标识别方法[J].激光与红外,2009,39(8):895-898. |
| 作者姓名： | 王琪 万中南 韩俊伟 |
| 作者单位： | 1. 中国兵器工业集团中兵光电科技股份有限公司,北京,100176;哈尔滨工业大学机电学院,黑龙江,哈尔滨,150001 2. 中国兵器工业集团中兵光电科技股份有限公司,北京,100176 3. 哈尔滨工业大学机电学院,黑龙江,哈尔滨,150001 |
| 基金项目： | 装备预先研究项目（No.102080301)资助 |
| 摘 要： | 单独使用中心矩进行目标识别，不仅算法耗时，而且效果也不理想。针对这个问题，提出了一种基于图像中心矩和特征向量相结合的目标识别方法，该方法具备了中心矩和可见光图像特征量各自的优点，降低了特征空间维数。通过引入匹配模板特征向量之间的欧式距离，进一步降低了算法的复杂度，提高了系统识别的速度，算例仿真验证了本方法的有效性和实时性。 |
| 关 键 词： | 特征提取 特征向量 欧式距离 模板匹配 |
Novel method of target recognition based on changeless rules and image feature vector |
|
| WANG Qi,WAN Zhong-nan,HAN Jun-wei.Novel method of target recognition based on changeless rules and image feature vector[J].Laser & Infrared,2009,39(8):895-898. | |
| Authors: | WANG Qi WAN Zhong-nan HAN Jun-wei |
| Affiliation: | 1.CNGC China North Optical-electrical Technology Co.;LTD;Beijing 100176;China;2.School of Mecha-tronics Engineering;Harbin Institute of Technology;Harbin 150001;China |
| Abstract: | An algorithm of target recognition using central changeless rules is not only time-consuming but also ineffective sometime.According to this problem,firstly a method of target recognition based on changeless rules and feature vectors is presented,this method has the advantages of both of them,so the characteristic special dimensions are reduced.Secondly the Euclidean distance among the feature vectors of matching templates is introduced,hence the algorithm complexity is decreased further,meanwhile the recognition velocity is increased.Finally the simulation proves that this method is effectual and efficient. |
| Keywords: | feature extraction feature vector euclidean distance template matching |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |
| 点击此处可从《激光与红外》浏览原始摘要信息 | |
| 点击此处可从《激光与红外》下载全文 | |